Abstract
Interfacial electronic band-matching (EBM) plays a crucial role in determining the spin-dependent transport properties and performance of spintronic devices. The final goal of this study is to establish a method to search for new material combinations that exhibit favorable EBM at the interfaces to achieve a superior performance in various spintronic devices using the machine learning technique combined with the first-principles calculations. As a first step, we investigate the effect of interfacial EBM on magnetoresistance (MR) by fabricating the current-in-plane giant magnetoresistive devices with compositionally graded Co1−βFeβ layers and Cu spacer. The MR ratio varies significantly across β = 0.11–1.0, with the highest MR of 17.5% observed at β ≈ 0.46, followed by a sharp decrease beyond β = 0.6. To analyze the β dependence of MR in terms of EBM with low computational cost, we calculate the simple Fermi surfaces of bcc Co1−βFeβ and Cu and evaluate the wave number (k) distance between their Fermi surfaces. The closest (furthest) Fermi surface match occurs at β ≈ 0.4 (1.0), which tends to be in good agreement with the observed MR trend. This suggests that a simple Fermi surface similarity analysis, when integrated with a machine learning technique, can be an effective method for efficiently identifying new material combinations with high EBM.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call